protests against war.
The argument for protection of unwilling audiences is much weaker where what is portrayed are mildly unconventional sexual attitudes or practices, not involving violence or domination. Here it is more plausible to say that discovering how one feels about such matters when accustomed to them is the best way of discovering what attitude towards them one has reason to hold. The lack of independent grounds for appraising these attitudes makes it harder to conceive of changes produced by expression as a kind of harm or corruption. Even here there are some independent grounds for appraisal, however. 32 Attitudes towards sex involve attitudes towards other people, and the reasons for or against holding these attitudes may be quite independent of one's reactions to portrayals of sex, which are, typically, highly impersonal. I believe that there are such grounds for regarding as undesirable changes in our attitudes towards sex produced by pornography, or for that matter by advertising, and for wanting to be able to avoid them. But, in addition to the problem of separability, just mentioned with regard to portrayals of violence, these grounds may be too close to the substantive issues in dispute to be an acceptable basis for the regulation of expression.
It seems, then, that an argument based on the need to protect unwilling audiences against being influenced could justify restriction of at most some forms of offensive expression. This leaves us with the residual question how much offense must be tolerated in order for persuasion and debate regarding sexual mores to go forward. Here the clearest arguments are by comparison with other categories of expression. The costs that audiences and bystanders are required to bear in order to provide for free political debate are generally quite high. These include very significant psychological costs, as the Skokie case indicates. Why should psychological costs of that particular kind occasioned by obscenity be treated differently (or given a particularly high value)? A low cost threshold would be understandable if the issues at stake were trivial ones, but by the would-be restrictors' own account this is not so. I do not find the prospect of increased exposure to offensive expression attractive, but it is difficult to construct a principled argument for restriction that is consistent with our policy towards other forms of expression and takes the most important arguments against pornography seriously.
*Versions of this paper were presented at the University of Minnesota and the University of California at Berkeley as well as at the Pittsburgh symposium. I am grateful to members of all these audiences for helpful comments. I have also benefited greatly from discussions of this topic with Marshall Cohen, Clark Glymour, and Derek Parfit.